FS and SeeYou

They gave slightly different distances for the tasks, and that's on purpose

I scored a couple of meets in Australia in order to test my SeeYou scoring system against the FS scoring system.  I was able to squash a couple of bugs in each program using these comparisons.

http://ozreport.com/12.067#2

Even though both programs gave the same competition results, I never did get the programs to give exactly the same distances for the tasks.  And the reason is that FS ignores the distance inside the turn point radius as seen here.

I did a quick check with the Great Circle calculator linked to above using the WGS 84 Earth model and kilometers.  Going from S 27.15887, E 151.26378 to S 26.77322, E 150.61653 I got 77.1720322871094 km.

SeeYou, for the same points, gave me 77.2 km.  FS gave me 76.7 km when these distances are used in a Task.  For FS take 77.2, subtract 400 meters for the radius of the first turnpoint and 100 meters for the small circle around the start point (ignoring the large start cylinder radius) and you get 76.7 km.

Stein-Tore Erdal <stein.tore.erdal> writes:

FS used the Haversine formula with an earth radius of 6371000 meters (FAI definition Sporting Code General Section 7.3.1.1).  In some cases (checking if a tracklog point is inside/outside a turnpoint cylinder) it converts coordinates to UTM and uses Pythagoras, which works fine for short distances.  FS also applies an error margin (default 0.5%) of turnpoint radius) when checking.  This should cover any difference in calculation there may be from the flight-instrument to FS.

In FS 1.2 (http://fs.fai.org/cgi-bin/trac.fcgi/wiki/Fs1.2) you can test distance calcs in FsFlight under Tools menu.

Next version of FS (1.2.1) will use the Haversine formula with FAI earth radius for all distance calculations.

The reason I used Pythagoras was that in some earlier version I had all the tracklog-points stored internally in UTM as well as degrees.  Then the computation is a lot faster with Pythagoras.  In current versions of FS the tracklog-points is only stored as degrees so now it is just as fast to use Haversine directly rather than converting to UTM then using Pythagoras.

Have done some checking with XC-Trainer and Garmin.  For a couple of Manilla waypoints going north/south, for example: 335KMS, S 27 45 19.74, E 149 44 45.96 to BAANBG S 30 35 58.80, E 149 59 05.82

My Garmin gives 316.1km when using go to from one to the other and looking in the map display.  316km when looking at route.  My XC-Trainer shows 317.1km.  FS gives 317.107km.  My Compeo+ is broken and sent to rep.  (could not handle the snow :-( )

So, for example, having a 300km radius start around 335KMS, I would look at my XC-Trainer while waiting for the start gate and being 1km closer than the other guys using Garmin.

I've done similar tests before and had around 150 meters difference between Garmin and XC-Trainer for 50km radius start.

Discuss FS and SeeYou at the Oz Report forum   link»